Bone marrow fat accumulation is part of dynamic processes that affect bone density and strength. The ability to measure marrow fat non-invasively has greatly advanced clinical research in this area. Cross-sectional studies have identified associations between higher marrow fat and lower bone density (BMD) measured by dual x-ray absorptiometry (DXA) and quantitative computed tomography (QCT). Importantly, previous research shows that marrow fat is associated with prevalent vertebral fractures, independent of BMD. These cross- sectional results are encouraging but cannot address key questions regarding the dynamic processes connecting marrow fat and bone. At this juncture, longitudinal studies are needed to clarify whether marrow fat predicts bone loss and to determine if changes in marrow fat are accompanied by changes in bone. Longitudinal studies are also needed to determine the role of marrow fat during weight loss, an important risk factor for bone loss and fracture risk in older adults. Severe calorie restriction increases marrow fat. Research is needed to determine if bone loss caused by weight loss in older adults is mediated in part by increases in marrow fat. In studies of marrow fat and bone, it is important to account for the potential effects of other fat depots, including total body fat, visceral fat and fatty infiltration of muscle, that are associated with bone density and fracture and, in some studies, with marrow fat. Another crucial issue is to determine whether the gender differences suggested in cross-sectional studies of marrow fat, bone and vertebral fracture are observed in longitudinal studies and if these differences are due to sex hormone levels. Associations between higher marrow fat and lower BMD appear to be stronger in women. However, the association between higher marrow fat and prevalent vertebral fracture may be stronger in men. To address the need for longitudinal data on marrow fat and bone, taking into account sex hormone levels, changes in weight and in other fat depots, we propose a study among 557 older adults in the Iceland AGES Reykjavik cohort that will measure changes in bone at the spine and hip, using QCT and DXA, and changes in vertebral marrow fat, measured by magnetic resonance spectroscopy (MRS), as well as visceral and subcutaneous abdominal fat and thigh muscle fat infiltration (measured by QCT) and total body soft tissue composition (measured by DXA), over a period of 2-5 years, with sex hormone levels and prevalent vertebral fractures measured at baseline. This project will determine if changes in marrow fat are associated with changes in BMD, independent of changes in other fat depots; if changes in marrow fat account for any of the association between weight loss and bone loss; if sex hormone levels alter the relationship between marrow fat and BMD; and if higher marrow fat is independently associated with vertebral fractures in both men and women. This study promises to substantially advance our understanding of the dynamic relationship between marrow fat and bone, and these advances may translate into new diagnostic and therapeutic approaches for prevention and treatment of osteoporosis.